The present disclosure relates to computer processing systems, and more particularly to the parallel data collection and recovery of a failing virtual processing system.
Physical processing systems (or computing hosts) may utilize virtual processing systems (or computing guests) to virtualize operating systems and/or applications. Such systems typically leverage hypervisor technology to create and manage the many aspects of virtualized computing. For the purposes of this disclosure, a “virtual processing system” can refer to the entire system providing such a virtualized environment or to a single virtualized computing guest thereby supported. Any component of these systems may reach an erroneous and unrecoverable state during runtime, often necessitating a system-wide failure. At failure-time, it is important to collect the current state of the processors and memory of the physical processing system to allow for the underlying problem's subsequent analysis. It is also important to commence a recovery procedure, which may include re-initializing the physical processing system's hardware components before restarting the virtual processing system.